Lots of thanks this course is outstanding and it helps a lot I just have one question: Is there anyone who is taking this class and could explain how in practice we could calculate the variance of noise signal in each measurement ? I mean how to find the var for w1(noise in the 1st measurement) and for w2(2nd measurement), w3,w4 .......etc ?!
he says treat every Xi as a result of constant input theta + some noise, so by total variance law, the observation Xi's variance is just caused by the noise.
In 7:44 , the third term of the blue function shows (x_i -θ ) ^2 , but in 7:59 , the function shows (θ - x_i)^2 ? I think they are not the same
apologize for my poor math skill, (x-y)^2 is the same as (y-x)^2
Lots of thanks this course is outstanding and it helps a lot
I just have one question: Is there anyone who is taking this class and could explain how in practice we could calculate the variance of noise signal in each measurement ?
I mean how to find the var for w1(noise in the 1st measurement) and for w2(2nd measurement), w3,w4 .......etc ?!
It seems difficult. In practice we may need to assume the observations are IID?
he says treat every Xi as a result of constant input theta + some noise, so by total variance law, the observation Xi's variance is just caused by the noise.